WorksPhysics and Astronomy
The Skyrme-Dilaton and Nucleon Observables
Abstract
We extend the Skyrme model by incorporating the -meson as a dilaton. The motivation is twofold: the -meson is the lightest scalar resonance in QCD and is therefore the natural extensions to the pions of the classical Skyrme model, while its interpretation as a pseudo-Goldstone boson of spontaneously broken scale symmetry is supported by recent studies of gravitational form factors and related developments. Although the -meson is a broad resonance, for static observables it can be treated as real particle since its line-shape is not expected to be essential. A combined fit to static nucleon observables yields close to unity, with substantial improvement over the original Skyrme model, while simultaneously reproducing the physical value of , the overall scale of the model.
Prepared for submission to JHEP
Introduction
Since its proposal [1, 2], its formulation in QCD [3, 4, 5] (reviewed in [6, 7]), and its connection to fermion spin [3, 8], the Skyrme model has found wide-ranging applications [9] to nucleon and nuclear properties [10, 11]. Its relation to the large- expansion provides a qualitative justification for treating the baryon as a classical soliton [3]. Nevertheless, the minimal Skyrme model has well-known phenomenological limitations: when confronted with nucleon observables, its natural scale, set by the pion decay constant, falls short by about thirty percent, and many observables deviate by a similar amount. It has been suggested that the large- structure alone cannot account for these deviations, and that improvement should come from adding further mesons [12], in line with the idea that the baryonic sector emerges from the mesonic one.
In this spirit, many authors have added the lowest-lying vector mesons, the and/or the , motivated by their phenomenological relevance in the vector-meson-dominance approach [13, 14, 15, 16, 17, 18, 19]. The lowest-lying resonance in QCD, however, is the -meson. It is qualitatively different from the pions and the vector mesons: it is very broad [20] and has been in and out of the Particle Data Group (PDG) listings over the years [21].1 For static (Euclidean) observables considered here, its effects can nevertheless be represented by an effective scalar degree of freedom with a characteristic mass , as used in fits to gravitational form factors [22, 23] (and references therein).
Independent motivation for treating the as a genuine dynamical state, rather than a mere unitarisation object, has accumulated recently from two directions. On the lattice, studies of gauge theories near the edge of the conformal window consistently reveal a light flavour-singlet scalar (see [24, 25, 26] for recent works). In parallel, analytic work built on the hypothesis that QCD itself lies close to an infrared fixed point has identified the with the dilaton, the (pseudo-)Goldstone boson of spontaneously broken scale symmetry [27, 28, 29, 30]. This identification has now been confronted directly with lattice data on gravitational form factors [22, 23]: the residues extracted from a -pole fit agree with the predictions of dilaton effective theory, lending concrete, quantitative support to the picture that the couples like a dilaton rather than like an ordinary phenomenological meson. This convergence of lattice and analytic evidence provides a symmetry-based motivation for introducing the into the Skyrme model on this footing.2
In this work we therefore extend the Skyrme model by incorporating the -meson as a dilaton.3 The resulting Skyrme-dilaton theory combines the nonlinear realisation of chiral symmetry with a compensator field for Weyl transformations. The construction is based on the leading-order dilaton chiral effective Lagrangian, supplemented by the four-derivative Skyrme term. The improvement term required by local Weyl invariance is of particular importance: it fixes the coupling of the dilaton to the energy-momentum tensor and, consequently, the long-distance contribution of the scalar field to the gravitational form factor. In this way the soliton solution provides a concrete realisation of the connection between a dilaton and the nucleon’s gravitational -form factor.
The paper is organised as follows. In section 2, the model is presented and the asymptotics of the -field are derived. The numerical solutions are confronted against static observables in section 3. The paper ends with summary and conclusions in section 4. Additional plots and computational details are deferred to appendices.
The Skyrme-dilaton model
In our model, the and the are the Goldstone bosons of spontaneously broken chiral and scale symmetry
They transform non-linearly under chiral and (local) Weyl transformations
Intuitively, the -field does not transform under Weyl transformation since it is an angular variable. Throughout, will sometimes be referred to as the dilaton field.
The relevant Skyrme-dilaton (SD) Lagrangian is the leading order dilaton Lagrangian in chiral perturbation theory (χPT) [30, 43]
augmented by the Skyrme-term which can be conveniently written in terms of the (Lie algebra-valued) Maurer-Cartan form . The pion mass term for mass-degenerate quarks is given by
and includes the -potential. Expanding to second order in uncovers its mass
where will both be discussed shortly. The term proportional to the identity in (4) implements the subtraction of the disconnected part (vacuum expectation value), thereby ensuring a finite energy solution. In the limit , which will be referred to as the Skyrme-limit, one recovers the standard Skyrme Lagrangian [5]. The Wess-Zumino-Witten (WZW) term, being a 5-form, vanishes for as it is a 3-manifold only, and is therefore absent.
The construction principle of the Lagrangian (3) is invariance under the two symmetries discussed above. The -field acts as a compensator to implement local Weyl invariance. Since the Skyrme term has four derivatives, there is no -field coupling to it. The term proportional to the Ricci-scalar (improvement term) ensures local Weyl invariance together with the dilaton kinetic term. It is crucial as it implements the Goldstone matrix element
which defines the dilaton decay constant, the order parameter of spontaneous scale symmetry breaking analogous to
being the order parameter of spontaneous chiral symmetry breaking.
The effective Lagrangian (3) relies on two results from matching to QCD in a fixed-point interpretation: the quark-mass anomalous dimension [29, 30] at the fixed point and the vanishing of the -function slope [30] (see also [44]). The former ensures that assumes and the latter avoids further trace anomaly matching, as parametrised in [41] for example.
The Lagrangian (3) introduces the four parameters where
are dimensionless variables with already defined in (5). The quantities and are experimentally well-defined whereas is a low energy constant to be determined indirectly.
To discuss , we first need to clarify a few points. In the real world, the -meson is unstable and has a sizeable width, which may raise the question of whether its inclusion as an explicit degree of freedom is justified. The key point is that we consider only static (Euclidean) quantities. For such observables, the can be effectively mimicked by a stable particle with a mass of (see [22] and references therein). From (5), we see that this corresponds to . Although is not known a priori, there are indications that it is not too far from unity. For a more comprehensive assessment of the as a dilaton across more channels, the inclusion of the strange-quark mass is presumably essential [27, 30]. However, for the observables considered here, which do not involve isolated -channels, we expect this effect to be less important. Our results support this expectation, with treated as a fit parameter.
We briefly digress to discuss Derrick’s theorem [45] in the context of the Skyrme model since it motivates the addition of the Skyrme-term and provides a powerful numerical crosscheck. Let us first neglect the pion mass and rescale units by . We then have
where are the two- and four-derivative terms respectively. The parameter governs the size of the soliton and stabilises when it is extremal , implying .
So, clearly if there is no Skyrme-term , then , and there is no finite energy solution. Adding pion mass, which corresponds to a zero-derivative term, modifies the relation to
which is a powerful sum rule satisfied by minimised solutions. In the Skyrme context, this is sometimes referred to as (an instance of) the virial theorem.
Next one needs to minimise the energy which can be obtained from the energy-momentum tensor (EMT) by metric variation ( denotes the metric determinant and is the Minkowski metric with mostly-minus signature). One obtains
with . The energy-density part is given by
where . In anticipation of the static ansatz, the static limit was assumed such that denotes a 3D Euclidean Laplacian (i.e. summation over indices assumes a Euclidean metric ).
Hedgehog ansatz and equation of motion
The associated Hamiltonian is obtained from and to make further progress we now commit to the famous hedgehog ansatz for the skyrmion
which intertwines space and isospin rotations in a non-trivial manner since is the unit -vector and are the (isospin) Pauli matrices. The profile function , sometimes referred to as the skyrmion, is solely dependent on the radius. The total energy is obtained by minimising
with mass density
where primes denote derivatives with respect to and hereafter
Moreover, we have silently redefined5
in order to work with a dimensionless variable . The scale assumes about half a Fermi, , for characteristic references values of and . It sets the overall scale similar to the lattice spacing in Monte Carlo simulations.
The equations of motion for follow from variation by
and
where
The expressions are consistent with the equation of motion in [43]. Note that since we are in the static case, minimisation of the energy is equivalent to solving the equation of motion.
Rigid rotator quantisation
The solutions and are spherically symmetric by the nature of their ansatz, but the nucleon is not as it has spin. This can partly be captured by quantisation of the rotational degrees of freedom [5], also known as moduli or collective coordinates, which we will review here for completeness and for introducing the role of the dilaton. The starting point is the observation that for any , forms a new solution, and if is taken to be time-dependent then one obtains
where
In the Skyrme-limit , this agrees with Eq. 4.64 [7] and the expression in [5] when taking conventions into account. Time-dependence allows for canonical quantisation and gives the following correction [5]
upon using that isospin and angular momentum are intertwined . The quantity is the mass after inclusion of the “rigid-rotator” quantum corrections. The expression in (23) makes it clear that correspond to the nucleon and the -baryon respectively:
We see that the correction to the nucleon mass is corresponding to a 8%-correction. The quantisation procedure affects many other observables, as we will see in § 3.
The overall scale of the corrections is still set by the pion decay constant, but they involve different powers of . This affects their large- behaviour, which is worth a digression (see Appendix § 6 for an extended discussion). It is well established that and , implying 6 . Consequently, , which amounts to relative suppression of of the quantum-versus-classical contribution.
Boundary conditions
The ordinary coupled differential equations of second order (18) and (19), require two boundary conditions for each field. For the Skyrme field , these are related to the baryon current
which is a 3-form and can be obtained by gauging the WZW-term with left-right sources such that . Other terms do not contribute since the pion and dilaton fields are of baryon number zero. The dilaton field does not alter since it is a conserved current of scaling dimension three and does not require powers of as compensators. The number
is an integer characterising the homotopy and corresponds to the baryon number. The nucleon has baryon number , and the conventional choice is
which settles the -boundary conditions but one may go further and investigate the asymptotic behaviour. Assuming that the dilaton field is constant at both boundaries (to be confirmed), the same analysis applies as in the Skyrme model: demanding finiteness towards the integration for (14) and by linearising for it follows that
where is unknown and is related to the Goldberger-Treiman relation [5], see the discussion in appendix § 5, as we will do in the dilaton case in the next section.
We may then turn to the dilaton field. At large distances, (28) implies that we may neglect and from (19) one finds the solution
with a priori unknown constants and , and the minus sign chosen in anticipation of the result. At short distances and thus (and let us use as well) for which the equation solves to
Now, since we have a -term in the -integral, finiteness demands and thus
where we recall that and an unknown constant.
Let us return to the question of whether the - (28) and -asymptotics are self-consistent. It seems so. At short distances the -term drops and we note that for , out of the remaining four terms the -terms cancel separately in (18). At large distances, constant -terms are leading as per (29) and therefore equivalent to the standard Skyrme analysis.
For the actual numerical calculation, concrete boundary conditions are needed. In order to obtain a non-trivial solution we must impose a finite value at infinity. We may choose which is natural if we assume that the -field vanishes at infinity. This is not arbitrary, since formally one can always rescale (cf. (3)) by
and this leads to an overall factor of in front of the action. Thus one may always absorb the value at infinity by other parameters. Moreover, finiteness of the energy automatically enforces , consistent with our short distance analysis (31). So, in summary we have
where we will see in the next section that is determined by the dilaton Goldberger-Treiman relation ensuring the conformal Ward identity.
The -asymptotics and the conformal Ward identity
The key to satisfying the conformal Ward identity and retaining a finite nucleon mass is that the gravitational -form factor is to assume the form (see appendix § 7.§ 3)
which is realised by the dilaton Goldberger-Treiman relation [43, 46, 47, 48]
The central point is that in the spontaneously broken phase, the Ward identities are still intact, as is the case in the chirally broken phase where the pion Goldberger-Treiman relation [49] ensures the chiral Ward identity. In the Skyrme model, governs the pion far-field [5] and we aim to do the same for the far-field in relation to . Our presentation is logically different but equivalent to [5], which is sketched in appendix § 5.
In order to carefully track factors, we consider it worthwhile to work in dimensions in this section. Now, the -cloud of the nucleon is determined by the interaction , giving the equation of motion
Going to the non-relativistic limit and assuming the nucleons to be a point source , as appropriate for the far-field, one gets with
which is solved by the standard Green’s function :
Temporarily focusing back on , the far-field reduces to
connecting to (33) and changing to our units (17) one gets
such that in (33) becomes
In the second equality, we have assumed the dilaton Goldberger-Treiman relation and in the last one expressed the in terms of (14), thereby ignoring the quantisation part. This last step is appropriate since the energy function is minimised without the quantisation part (see the discussion in [50]). We will see that the numerical simulation reproduces this asymptotic behaviour precisely for towards zero.
As a next step we will compute the residue (34) from the gravitational form factor, using the solution in (40). To do this, we need an expression of the residue in terms of which is expressible in terms of the - and -fields. This can be achieved by considering the form factor in the Breit-frame (, ) where one has (see Eq. 17b [51])
Contracting indices and using (relativistic state normalisation) and the identification one gets the desired formula
This equation looks puzzling at first sight, as it is telling us that the theory has a (massless) dilaton unless the integral on the right-hand side vanishes. Hence, there must be such a condition in the generic case. This relation is known as the von Laue condition (see the review [51])
which states that the pressure integrated over the volume must vanish.7 The derivation uses the conservation of the energy-momentum tensor and integration by parts. In the dilaton case, it is the latter assumption which breaks down.
The key quantity must be the improvement term as it is the case in the momentum space derivation in leading-order perturbation theory [43]. The latter is modified in dimensions to
where is the naturally normalised scalar with dimension , and is the improvement coefficient. Then we find,
upon using , the equation of motion (37) for and the dilaton Goldberger-Treiman relation (35). This expression modifies the von Laue condition to a dilaton version
Finally with (43) we find the residue (34)
ensuring IR-conformality for .
Formally, the situation is analogous to classical electrodynamics, where plays the role of the scalar potential in the Poisson equation for a point charge distribution. In our case, the role of the point charge is taken on by the nucleon at far distance.
Observables, results and fits
We consider nine observables in the fit, including the previously discussed , , and , together with one magnetic moment and five radii obtained from the electromagnetic and weak form factors, see appendix C.1 and C.2. The experimental input is given in the respective tables.
Static nucleon observables
We briefly summarise all observables in terms of seven integrals, denoted by , which are defined in appendix D, Eq. (D.1).
Before discussing the nucleon mass, an important issue concerning the large- corrections from rigid-rotator quantisation needs to be addressed. Since, as previously stated, the soliton profile is obtained by minimising the Lagrangian without the rigid-rotator contribution, adding the latter to the nucleon mass violates the von Laue condition. See Appendix C of [50] for an excellent discussion of this issue.8 Following the consistent prescription advocated in [50], we retain only the leading-order large- expression. This amounts to dropping the rigid-rotator term in (24) and considering the nucleon– mass splitting rather than the mass itself
From the electromagnetic form factors we may define the four electromagnetic radii defined from the slopes of the electromagnetic form factors, separating into isospin and ,
and the two magnetic moments
Note that in the Skyrme model the radii are degenerate , in slight tension with experiment as can be seen from Table 1. A useful conversion factor is
| Obs () | Exp | Sky-dil | Sky | ||
| 92.22(14) | 84.7 | 70.4 | |||
| 939.6 | 961.9 | 2 | 1082.5 | 15 | |
| 288.8 | 315.7 | 8 | |||
| 0.7690(13) | 0.68 | 0.57 | |||
| 0.82(9) | 0.86 | 5 | 0.82 | 0 | |
| 0.9068(11) | 0.91 | 1 | 1.05 | 16 | |
| 0.856(16) | 0.91 | 7 | 1.04 | 23 | |
| 0.8798 | 0.74 | 0.62 | |||
| 0.559(33) | 0.51 | 0.34 | |||
| 0.91 | 6.08 |
Table 1. The fit values are and for the Skyrme-dilaton, and for the Skyrme. All values are from the PDG [20] except the FLAG average [55] of the pion decay constant and the axial radius is taken from [56]. The error on the nucleon mass and are negligible. Whereas the is unstable with a pole of , we do not assign an extra uncertainty as this will be covered by the generic 10% large- model uncertainty assigned.
where for reference values of and . From the weak form factor , we get the nucleon axial charge and the nucleon axial radius (C.6)
The asymptotic behaviour of [5], see the discussion in appendix A, is given by
which upon using the pion Goldberger-Treiman relation and the -expression above assumes the form
which is exactly satisfied in the zero-pion-mass limit.
Comparing the observables obtained in the Skyrme model to real-world physical observables is not straightforward. The latter are at fixed , whereas the Skyrme model misses the -suppressed pion loops. Perfect agreement should therefore not be expected, and a corresponding large- uncertainty must be assigned. For definiteness, we take this uncertainty to be 10% for all observables.
Since we intend to fit the three parameters , and , we must identify observables that receive corrections well beyond this expectation. For the quantities , this has already been done in the pioneering paper [52], but in recent years more and more is known on large from lattice QCD (see Refs. [53, 54] for reviews). For other observables, this is assessed in appendix B where it is found that and (55) have large corrections in and are therefore not suitable for a fit since they would completely bias the fit-procedure due to the large corrections.
Numerical solutions
The numerical solutions are found by minimising the energy functional, which is equivalent to solving the equation of motion for the static case. The strategy for finding the solutions is described in appendix E.2 alongside plots with the Skyrme and the dilaton profile functions in Figure 2 for various values of . For illustrative purposes, a basic plot for , comparing the Skyrme-dilaton and the Skyrme model, is shown in Figure 1. There are two non-trivial test for the solutions.
Figure 1. Plot showing the dilaton function (solid black) as a function of in our units of (such that 2 units correspond to the proton charge radius). We also show the Skyrme-dilaton function (solid purple) versus the Skyrme function (purple dashed) for physical and (purple dotted) for close to the best fit-value in that model. The value with has been chosen for the Skyrme-dilaton model and is set to its physical value.
Figure 2. Plots of dilaton and Skyrme as a function of in our units of (top) physical pion mass (bottom) massless pion, for chosen values of implying .
a) The Skyrme virial theorem (10), is satisfied within numerical precision. b) The asymptotic form (40), in particular given in (41). For (which here implies ) where conformal symmetry appears unbroken, disregarding potential changes due to quantum corrections, we expect to be matched exactly and this is indeed the case to within numerical accuracy. For the dilaton Goldberger-Treiman relation ceases to be exact, in particular because (5) and there are corrections which are modest for low -masses that is large, as can be inferred from Table 3.
| 2.0 | 1.8 | 1.6 | 1.4 | 1.2 | 1.0 | |
| 1.00 | 1.11 | 1.25 | 1.42 | 1.67 | 2.00 | |
| 1.2 | 1.5 | 1.9 | 2.6 | 3.7 | 5.7 | |
| 1.17 | 1.41 | 1.72 | 2.15 | 2.78 | 3.78 | |
| 0.98 | 0.94 | 0.91 | 0.83 | 0.75 | 0.66 |
Table 3. Above . The quantity is the one given in (40) as extracted from the asymptotic form which equals the prediction in (41) to the given numerical precision. For the non-zero (2.5) leads to corrections in the dilaton Goldberger–Treiman relation (2.35)` as given in the last row.
| Obs () | Exp | Sky-dil | Sky | ||
| 80.3(3) | 87.1 | 8 | 63.16 | ||
| 836.24 | 824.3 | 884.4 | 6 | ||
| 260.4 | 268.4 | 3 | 254.4 | ||
| 0.7690(13) | 0.75 | 0.63 | |||
| 0.82(9) | 1.0 | 23 | 0.99 | 20 | |
| 0.88 | 0.72 | 0.52 | |||
| 0.559(33) | 0.48 | ||||
| 2.18 | 5.69 |
Table 2. See Tab. 1, and text for which observables have been corrected for . The fit values are and for the Skyrme-dilaton, and for the Skyrme.
2-fits to nucleon observables
In the Skyrme-dilaton model, there are four parameters, , as introduced in Eq. (8). We fix
and fit and for a range of values of , while imposing the physical pion mass, MeV [20]. This parametrisation is convenient because the equations of motion must be solved separately for each fixed pair . Since is empirically expected to be of order unity ([30] and references therein), we have scanned the range .
The standard is defined by
where is the array of observables given in Table 1. The uncertainties entering the fit are encoded in the covariance matrix . For the -baryon which is unstable the-width-mass-ratio is ( MeV) but since the width is large- suppressed we do not assign an additional uncertainty to it.
In addition to the experimental uncertainties, we must account for the theoretical uncertainties associated with the large- approximation and the incomplete treatment of strange-quark mass effects (linked to the small mass as previously discussed). Since our strategy is to exclude observables with systematically large corrections, such as , and the pion Goldberger–Treiman asymptotics, from the fit, we assign a total 10%-systematic uncertainty as we simply cannot expect results to be better than that. This is in line with the long-standing rule of thumb for well-behaved large- observables in the literature [12].
The experimental and theoretical uncertainties are then added in quadrature to construct the covariance matrix.
The Skyrme-dilaton model
For physical pion mass, with the procedure described above, we get
a close to one, resulting in a -value of , with close to its physical value of MeV. The value of which is a parameter of the effective theory is slightly lower than for the original Skyrme model.
It is also interesting to consider the massless-pion limit and assess whether the resulting physics remains credible. In the real world we expect most observables to be smooth in the chiral limit, exceptions are the electromagnetic charge radii which diverge and therefore must be discarded (and the axial radius has been found to be unstable and is discarded for the time being). For the remaining observables (and in particular ) therefore constitute valuable testing ground. For the uncertainty analysis, it would be desirable to know all observables in the limit. In particular, since the strange-quark mass is mainly modelled through , which nevertheless vanishes as , we must consider the limit since the strange quark is expected to contribute significantly to the -mass in the dilaton case. The pion decay constant and nucleon mass are among the observables that are well studied in this limit: MeV [57] (in section 5.3.5), while is known to decrease by about 11% (see the review [58]). We assume the same reduction for the -baryon. For the remaining three quantities we are not aware of chiral estimates of the same reliability, but we may simply add 10% uncertainty in quadrature to the 10% which then raises to systematic uncertainty.
Taking these changes into account and dropping the two infinite radii, we get
which has a -value of . This is somewhat lower than in the massive pion case, as one would expect on grounds of not adapting the observables to the chiral limit. It is remarkable that the model does not lose its ground when the pion and the are made massless. Testing this matter was one of the main aims of this work.
Let us return to the two observables, and , which were omitted from the fit based on large- considerations. For the fitted values of , they take the values and [] in the massive and massless cases, respectively. While the values of are reasonably close to experiment and including it would not significantly alter the fit, the same is not true for .
Another aspect worthwhile emphasising is that the asymptotic behaviour due to the classical analysis, predicted in (33) and (41),
with , holds exactly within numerical precision for the computer solution.
The Skyrme model
In order to put these results into perspective, it is useful to assess the standard Skyrme-model () limit equally. We find
corresponding to a considerably larger (corresponding to ) than the ones obtained in the Skyrme-dilaton model. For the massive and massless case we have nine and seven observables respectively and the number of degrees of freedom are therefore seven and five respectively since we minimise over and . The values of and are obtained in the massive and massless case, respectively. The latter are not too far off from experiment and including them into the fit does, unlike , not change the overall picture of the fit considerably. The slightly different values than in the fits [5, 52] are due to the inclusion of the axial radius and the exclusion of .
Summary and conclusions
We have extended the Skyrme model by incorporating the as a dilaton. The motivation for doing so is both phenomenological and theoretical. The -meson is the lightest scalar resonance in QCD and therefore provides a natural additional low-energy degree of freedom, while its possible interpretation as a pseudo-Goldstone boson of spontaneously broken scale symmetry gives a symmetry principle governing its couplings. Since we have restricted our analysis to static quantities, the broad width of the physical can be represented by an effective stable scalar.
We solved the coupled equations for the pion and dilaton profiles and confronted the resulting model with nine static nucleon observables. The parameters were determined from a combined -fit, with uncertainties enlarged to account for the expected limitations associated with the large- nature of the Skyrme model. For a physical pion mass and 9 we obtain (59)
which is the main result of this paper. The quality of the fit is therefore substantially better than for the minimal Skyrme model, for which we find . In particular, the fitted value of is considerably closer to its physical value which was the main challenge for the original Skyrme model, see however appendix B for a more nuanced assessment. The improvement indicates that the plays a significant phenomenological role and at this level is consistent with coupling like a dilaton.
Besides investigating the real-world impact of the coupled as a dilaton on physical observables, one might wonder what happens in the more extreme case of a massless dilaton. Is it in principle possible that the -meson becomes massless in the chiral limit? After adapting some observables to this limit, and discarding the ones that diverge, we obtained (60) corresponding to a -value of approximately . This is considerably lower than the fit at the physical pion mass. The result should, however, not be overinterpreted: in the present effective model the simultaneous limit also drives , while the treatment of the strange-quark mass and corresponding observables is necessarily approximate. The limit demonstrates that it is not impossible that the -meson becomes massless in the chiral limit.
A theoretical result is the derivation of the dilaton far-field asymptotics (40), linked to the dilaton Goldberger–Treiman relation. For it is exact and verified by our numerical solution whereas for there are controlled deviations due to not being zero, see Table 3. Another curious aspect is the amendment of the von Laue condition (47) to the dilaton case.
Finally, our results suggest that the dilaton provides a useful extension of the Skyrme model and, importantly, a concrete framework in which the connection between scale-symmetry breaking and the -meson can be investigated. More broadly, our findings lend support to the possibility that the -meson couples approximately as a dilaton, the pseudo-Goldstone boson of spontaneously broken scale symmetry, and are consistent with the picture of QCD approaching an infrared fixed point. How far the -pole moves towards the infrared in the chiral limit remains an open question. This can be further explored by scrutinising the model against Euclidean form-factor data and by incorporating more realistic effects of the strange-quark mass. Further progress on the axial nucleon charge would also be desirable, either from reliable large- estimates or by incorporating such effects into the Skyrme-dilaton approach.
Acknowledgements: We are grateful to Christoph Adam, Mohamed Anber, Claudio Bonanno, Cedric Lorcé, Valya Khoze, Donal O’Connell, the late Mannque Rho, Bernd Schroers, Paul Sutcliffe and Tom Winyard for discussions. RZ is supported by STFC via the consolidated grants ST/T000600/1 and ST/X000494/1.
Note added: While completing this work, we became aware of related work [59] in which the -meson is coupled to the BPS Skyrmion [60]. This approach shares certain features with, but also differs in important respects from, the dilaton model considered here.
Connection of the - to the -asymptotics in ANW
In the pioneering work [5], a connection is made between the asymptotic behaviour of the pion field and the pion Goldberger–Treiman relation, analogous to the discussion presented in § 2.4 for the dilaton case. The presentation, however, differs somewhat, and we therefore consider it worthwhile to parallel the two derivations.
For that purpose it is useful to introduce the auxiliary form factor
such that is the analogue of . In the massless dilaton case we have (we work in the limit in this section). Their presentation in section 4 [5] proceeds in three steps:
Asymptotic value from numerical simulation: we know that the far-field is given by
and that for conserved currents () only the asymptotic form will play a role for the zero-momentum-transfer form factor. Hence, we may take the asymptotic ansatz, insert it into (A.2), and obtain expressions for and , respectively. We can then take and from the numerical solution to confirm the same value as for the zero-momentum form factor. This is a non-trivial tests of the formalism and the numerics.
Goldberger-Treiman relation: one may do so by solving for the asymptotic fields classically as done in (37), and obtain
where here has mass dimension of minus one (not rescaled as in (17)) and by choice see the discussion around (32)). The next step is to relate and to and as follows:
which reveals the pion and dilaton Goldberger-Treiman relation which guarantee that the chiral and conformal Ward identities are satisfied. We reemphasise that the entering the equation is only the classical part of the nucleon mass since we minimise with respect to the classical part.
At last it seems worthwhile to contrast the two asymptotic formulae for . In this case they become
where the form for is given in [52] and the second one has been derived earlier (40) (in units where is dimensionless though).
Large considerations
As the Skyrme model operates in the large limit it is important to assess which observables are reasonably accessible in such a limit. It is often said that the large limit works to 10% which we take as a guideline for the discussion below and the uncertainty estimate.
A point raised in [5] is that there are more observables than integrals, here (D.1), and this means that not all observables are independent within the model. In [5] relations are given for the magnetic moments . Whereas the one for is well-satisfied in nature, the one for is not. Hence, no solution can improve this matter, but rather one has to change the model more substantially by adding vector mesons for example as pointed out in [5]. We therefore drop from the fit and quote its value independently.
A particularly delicate quantity is the weak axial charge, whose sensitivity has been discussed from various perspectives [61]. For example, in the naive quark model one finds , suggesting potentially sizeable corrections unless significant cancellations occur coming from other contributions. We will therefore not consider it for the fit as otherwise it will completely bias the resulting parameters. Since we will find that and have similar large corrections, the Goldberger-Treiman relation then implies that has corrections of the same order as and we can therefore not expect it to be well-behaved either.
More promising observables are the radii, which are determined by the derivative of the form factor normalised to its value at zero momentum transfer. The large emergent spin-flavour symmetry [62, 63, 64], related to , assures the electric and magnetic form factors have relative corrections of
such that and are therefore in the class of observables that are well-behaved in the large- limit. This includes the axial radius.
While the integrals in (D.1) are, of course, independent of , our parameters and are not. Their -dependence must therefore also be taken into account, as it may distort the picture. A lot of knowledge has been accumulated in the last decade and it is found that many observables have rather sizeable large- corrections. For example, using that , one may define the appropriate quantity
it is found that the pion decay constant is reduced by about 30% in the linear fit [65, 66] which makes use of finite lattice data [67]. At first thought this calls into the question one of the main results here namely that is fitted to be close to its physical value. In addressing this question we must also address other observables such as the nucleon mass, since . Defining similarly, taking into account that , one infers from Fig. 12 in [54] (based on [68, 69]) that the nucleon mass also reduces considerably .10 If our result is to be meaningful, the change in has to be moderate.
We are not aware of a direct study of the large- behaviour of , and therefore take a detour by employing two phenomenologically successful relations: the KSRF relation, , and the hidden local symmetry relation, . Together, these give which is reasonably close to the values obtained in Skyrme-model studies.11 For our purposes, the main advantage of this connection is that it allows us to infer the large- behaviour of from that of . Combining with the large- correction of the string tension, [69], which exhibits only a mild dependence, we obtain . This finally yields which we use as a reference value.12
This provides further consistency checks, in particular as to whether imposing the physical values of and is meaningful within the Skyrme model. For this to be the case, the following ratio should be close to unity: , which is indeed satisfied. Interestingly, this is also the combination that ensures that remains a well-behaved observable, as can be seen from (51). The mass splitting scales as , and the corresponding quantity, , is not small. Nevertheless, since we assign an additional uncertainty to this quantity to account for the finite width, we retain it in our analysis. Finally, all radii, being dimensionful quantities, scale as . However, since , as found above, the corresponding corrections remain within a tolerable range.
In conclusion, we may keep all quantities but , and must be cautious with the asymptotic condition due to the pion Goldberger–Treiman relation.
Definition of observables
For completeness, we define the observables included in our fit, which are derived from the electromagnetic, weak and gravitational form factors in sections C.1, C.2 and C.3, respectively.
Electromagnetic form factors
The electromagnetic Dirac and Pauli form factors are defined by (, )
and related to the Sachs form factors by
Their values at zero momentum transfers and their slopes define prominent nucleon observables. For example, the proton and neutron magnetic moments
where is the nucleon magneton and is the electric charge (charges are trivial and ). In the Skyrme model it is customary to define the electromagnetic form factors without electric charge and thus is dropped from . Moreover, because of isospin symmetry of QCD (for degenerate up and down quark masses), it is convenient to define the isospin combinations
The slopes define the four radii
from where we may define the isospin combinations
and
Weak form factors
The weak form factors are defined by
for with an flavour matrix. The weak charge and the so-called axial radius13 are defined by
Since we are to use the Goldberger-Treiman relation to compute , it seems worthwhile to sketch how it derives from Eq. (C.5). First, one writes
with the pion pole appearing in the form factor, and can be defined from the effective Lagrangian
Then, one observes that for the chiral Ward identity to be obeyed, the pion Goldberger-Treiman relation [49]
must hold for massless pions.
Gravitational form factors
It is convenient to define , from where the three form factors follow
with
Lorentz structures ensuring translational invariance . The normalisation holds model-independently, since is the associated Noether current with conserved momenta . Furthermore, for the nucleon one has , which reflects the absence of an anomalous gravitational magnetic moment, a consequence of the universality of gravity. The form factors and are analogues of the electromagnetic, Dirac and Pauli form factors, given in (C.1). Contracting one gets
and we see that () if all form factors are regular in the limit but for a massless dilaton: .
The seven basic integrals entering the observables
For convenience, we reproduce all integrals entering the observables (using (2.16))
where the formulae of the two axial observables are sensitive to whether the pion is massive or not, as explained in section D.1 below. The remaining cases can be obtained from the literature by multiplying the kinetic contributions by a factor of , with some additional modifications for , which are discussed in detail in the main text. When comparing with [5], one should also account for the different convention used for , as explained in footnote 4. The expression for is unchanged since the dilaton does not couple to the baryon current and is easily computed from (41) and (13); it is given below Eq. 18 in [5]. The same applies to , which has a dilaton addition in the term proportional to in [5]. The expression for agrees with Eq. 4.64 in [7] and can be shown to be equivalent to the expression given in [52] upon taking the several conventions into account. The integral for , after adding the -factor, is equal to Eq. 4.105 [7], correcting an overall sign mistake, and equivalent to Eq. 30 in [5] after taking conventions into account.
The axial observables
The axial form factor (and ) have been derived in [72] (Eq. 4.17). In the massless pion case, the axial Ward identity allows for simplifications and are at the origin of case distinctions in (D.1). The generally valid result for the axial form factor reads
where
with and and
adapted to the dilaton case and taking into account their conventions. In the massless pion case, the axial current is conserved, as a matrix element, implying relations amongst the axial and pseudoscalar form factor which amount to
and this explains the two axial formulae in (D.1). We have checked that the formulae in (D.1) are numerically continuous in the limit.
Plots and numerics
Plots of Skyrme-dilaton solution for variable
Plots of the Skyrme-dilaton solution are given in Figure 2 for variable for the massless and the massive pion case.
Finding classical field configurations numerically
While the equations of motion Eqs. (18) and (19) can be solved numerically via a shooting approach, this is generally not a promising strategy. If one were to start at , radially integrate outward and then try to use some optimization strategy to satisfy boundary conditions at , this would run into the same problem that we would encounter when trying to find energy eigenstates of the harmonic oscillator in such a way: there are different viable options for large- behavior that correspond to different solutions. For the harmonic oscillator, these differ in eigenstate energy, while here, they differ in Baryon number. Being able to “jump” between such solutions tends to confuse whatever numerical search strategy is employed (which may well include elements such as penalty terms for trial solutions that depend on total energy) in conjunction with a shooting method.
A more powerful – but in terms of numerical algorithms still rather basic – approach is to instead use the Hamiltonian to work out the energy of a classical configuration, discretise the problem by picking sampling points at which the value of and are used as the degrees of freedom of the discretised approximation to the continuum problem, then obtain and via algorithmic differentiation of some (e.g. cubic) spline interpolation with continuous first derivatives and use that data alongside the original interpolants to numerically estimate the total energy of the candidate field configuration. Here, care has to be taken to employ an integration scheme for energy density that is compatible with – or ideally directly derived from – spline interpolation, for one can not expect high-order Gauss-Kronrod quadrature to produce meaningful results if interpolation order is truncated at some low polynomial degree. It furthermore makes sense to transform the integration range to some finite interval by a transformation such as .
Numerical minimisation then yields useful approximations to the desired ground state if one pins and in order to enforce obtaining a solution, while also pinning . Even when not enforcing a fourth boundary condition, (finite) energy minimisation produces the desired solution.
The physical idea is that the value of the dilaton field approaches some constant at infinity, and we can make that constant equal to 1 by suitable redefinition of . It is useful and rather manageable even on modest computing hardware to pick around sampling points and employ a higher order nonlinear optimization algorithm such as conjugate gradient or BFGS [73] minimisation, using gradients that most conveniently can be obtained via automatic differentiation.
This basic approach is able to produce a high-quality solution with 1000 support points on a somewhat modern laptop in less than 30 minutes.
References
References
- [1]T. H. R. Skyrme, “A Nonlinear field theory,” Proc. Roy. Soc. Lond. A 260 (1961) 127–138.
- [2]T. H. R. Skyrme, “A Unified Field Theory of Mesons and Baryons,” Nucl. Phys. 31 (1962) 556–569.DOI
- [3]E. Witten, “Baryons in the 1/n Expansion,” Nucl. Phys. B 160 (1979) 57–115.
- [4]A. P. Balachandran, V. P. Nair, S. G. Rajeev, and A. Stern, “Exotic Levels from Topology in the QCD Effective Lagrangian,” Phys. Rev. Lett. 49 (1982) 1124. [Erratum: Phys.Rev.Lett. 50, 1630 (1983)].
- [5]G. S. Adkins, C. R. Nappi, and E. Witten, “Static Properties of Nucleons in the Skyrme Model,” Nucl. Phys. B 228 (1983) 552.
- [6]J. F. Donoghue, E. Golowich, and B. R. Holstein, Dynamics of the standard model, vol. 2. CUP, 2014.
- [7]Y.-L. Ma and M. Harada, “Lecture notes on the Skyrme model,” arXiv:1604.04850 [hep-ph].arxiv.org/abs/1604.04850
- [8]D. Finkelstein and J. Rubinstein, “Connection between spin, statistics, and kinks,” J. Math. Phys. 9 (1968) 1762–1779.
- [9]G. E. Brown and M. Rho, eds., The multifaceted skyrmion. World Scientific, Singapore, 2010.arxiv.org/abs/0907.1963
- [10]N. S. Manton and P. Sutcliffe, Topological solitons. Cambridge Monographs on Mathematical Physics. Cambridge University Press, 2004.DOI
- [11]N. S. Manton, Skyrmions – A Theory of Nuclei. World Scientific, 3, 2022.DOI
- [12]E. Witten, “SKYRMIONS AND QCD,” in Nuclear Chromodynamics: Quarks and Gluons in Particles and Nuclei. 6, 1984.
- [13]G. S. Adkins and C. R. Nappi, “Stabilization of Chiral Solitons via Vector Mesons,” Phys. Lett. B 137 (1984) 251–256.
- [14]U. G. Meissner and I. Zahed, “Skyrmions in the Presence of Vector Mesons,” Phys. Rev. Lett. 56 (1986) 1035.pubmed.ncbi.nlm.nih.gov/10032551
- [15]U. G. Meissner, N. Kaiser, A. Wirzba, and W. Weise, “Skyrmions With ρ and ω Mesons as Dynamical Gauge Bosons,” Phys. Rev. Lett. 57 (1986) 1676.DOI
- [16]U. G. Meissner, N. Kaiser, and W. Weise, “Nucleons as Skyrme Solitons with Vector Mesons: Electromagnetic and Axial Properties,” Nucl. Phys. A 466 (1987) 685–723.DOI
- [17]P. Sutcliffe, “Multi-Skyrmions with Vector Mesons,” Phys. Rev. D 79 (2009) 085014, arXiv:0810.5444 [hep-th].DOI
- [18]Y.-L. Ma, Y. Oh, G.-S. Yang, M. Harada, H. K. Lee, B.-Y. Park, and M. Rho, “Hidden Local Symmetry and Infinite Tower of Vector Mesons for Baryons,” Phys. Rev. D 86 (2012) 074025, arXiv:1206.5460 [hep-ph].DOI
- [19]C. Naya and P. Sutcliffe, “Skyrmions in models with pions and rho mesons,” JHEP 05 (2018) 174, arXiv:1803.06098 [hep-th].DOI
- [20]Particle Data Group Collaboration, F. Takahashi et al., “Review of Particle Physics,” Int. J. Mod. Phys. A 41 no. 22, (2026) 2630011.
- [21]J. R. Pelaez, “From controversy to precision on the sigma meson: a review on the status of the non-ordinary f0(500) resonance,” Phys. Rept. 658 (2016) 1, arXiv:1510.00653 [hep-ph].
- [22]R. Stegeman and R. Zwicky, “Gravitational D-form factor: the σ-meson as a dilaton confronted with lattice QCD data I,” JHEP 03 (2026) 184, arXiv:2508.18537 [hep-ph].
- [23]R. Stegeman and R. Zwicky, “Gluon Gravitational D-Form Factor: The σ-Meson as a Dilaton Confronted with Lattice Data II,” arXiv:2512.12315 [hep-ph].arxiv.org/abs/2512.12315
- [24]Z. Fodor, K. Holland, J. Kuti, D. Nogradi, and C. H. Wong, “Extended investigation of the twelve-flavor β-function,” Phys. Lett. B 779 (2018) 230–236, arXiv:1710.09262 [hep-lat].
- [25]Lattice Strong Dynamics Collaboration, R. C. Brower et al., “Light Scalar Meson and Decay Constant in SU(3) Gauge Theory with Eight Dynamical Flavors,” arXiv:2306.06095 [hep-lat].
- [26]LatKMI Collaboration, Y. Aoki, T. Aoyama, E. Bennett, T. Maskawa, K. Miura, H. Ohki, E. Rinaldi, A. Shibata, K. Yamawaki, and T. Yamazaki, “A novel view of the flavor-singlet spectrum from multi-flavor QCD on the lattice,” arXiv:2505.08658 [hep-lat].
- [27]R. J. Crewther and L. C. Tunstall, “ΔI = 1/2 rule for kaon decays derived from QCD infrared fixed point,” Phys. Rev. D 91 no. 3, (2015) 034016, arXiv:1312.3319 [hep-ph].
- [28]R. J. Crewther and L. C. Tunstall, “Status of Chiral-Scale Perturbation Theory,” PoS CD15 (2015) 132, arXiv:1510.01322 [hep-ph].arxiv.org/abs/1510.01322
- [29]R. Zwicky, “QCD with an infrared fixed point: The pion sector,” Phys. Rev. D 109 no. 3, (2024) 034009, arXiv:2306.06752 [hep-ph].arxiv.org/abs/2306.06752
- [30]R. Zwicky, “QCD with an infrared fixed point and a dilaton,” Phys. Rev. D 110 no. 1, (2024) 014048, arXiv:2312.13761 [hep-ph]. [Erratum: Phys.Rev.D 113, 039902 (2026)].arxiv.org/abs/2312.13761
- [31]G. E. Brown and M. Rho, “Scaling effective Lagrangians in a dense medium,” Phys. Rev. Lett. 66 (1991) 2720–2723.pubmed.ncbi.nlm.nih.gov/10043599
- [32]M. Rho and Y.-L. Ma, “Manifestation of Hidden Symmetries in Baryonic Matter: From Finite Nuclei to Neutron Stars,” Mod. Phys. Lett. A 36 no. 13, (2021) 2130012, arXiv:2101.07121 [nucl-th].DOI
- [33]M. Rho and L.-Q. Shao, “Probing for an IR-fixed Point in QCD by Superallowed Gamow-Teller Transitions in Doubly Magic Nuclei,” arXiv:2410.04991 [nucl-th].arxiv.org/abs/2410.04991
- [34]D. Elander and M. Piai, “Calculable mass hierarchies and a light dilaton from gravity duals,” Phys. Lett. B 772 (2017) 110–114, arXiv:1703.09205 [hep-th].DOI
- [35]D. Elander and M. Piai, “Glueballs on the Baryonic Branch of Klebanov-Strassler: dimensional deconstruction and a light scalar particle,” JHEP 06 (2017) 003, arXiv:1703.10158 [hep-th].DOI
- [36]D. Elander, M. Piai, and J. Roughley, “Dilatonic states near holographic phase transitions,” Phys. Rev. D 103 (2021) 106018, arXiv:2010.04100 [hep-th].DOI
- [37]A. F. Faedo, C. Hoyos, M. Piai, R. Rodgers, and J. G. Subils, “Light holographic dilatons near critical points,” Phys. Rev. D 110 no. 12, (2024) 126017, arXiv:2406.04974 [hep-th].arxiv.org/abs/2406.04974
- [38]D. Elander, A. F. Faedo, M. Piai, R. Rodgers, and J. G. Subils, “Light dilaton near critical points in top-down holography,” Phys. Rev. D 112 no. 12, 126020, arXiv:2502.19226 [hep-th].
- [39]S. Matsuzaki and K. Yamawaki, “Holographic techni-dilaton at 125 GeV,” Phys. Rev. D 86 (2012) 115004, arXiv:1209.2017 [hep-ph].DOI
- [40]D. D. Dietrich, F. Sannino, and K. Tuominen, “Light composite Higgs from higher representations versus electroweak precision measurements: Predictions for CERN LHC,” Phys. Rev. D 72 (2005) 055001, arXiv:hep-ph/0505059.DOI
- [41]O. Catà, R. J. Crewther, and L. C. Tunstall, “Crawling technicolor,” Phys. Rev. D 100 no. 9, (2019) 095007, arXiv:1803.08513 [hep-ph].DOI
- [42]M. Tanaka, D. Fujii, and M. Kawaguchi, “Gravitational form factors of the nucleon in the Skyrme model based on scale-invariant chiral perturbation theory,” arXiv:2507.21220 [hep-ph].DOI
- [43]R. Zwicky, “Dilatons improve (non)-Goldstones,” Nucl. Phys. B 1023 (2026) 117292, arXiv:2306.12914 [hep-th].DOI
- [44]M. Shifman and R. Zwicky, “Relating β* and γQ* in the N=1 SQCD conformal window,” Phys. Rev. D 108 no. 11, (2023) 114013, arXiv:2310.16449 [hep-th].arxiv.org/abs/2310.16449
- [45]G. H. Derrick, “Comments on nonlinear wave equations as models for elementary particles,” J. Math. Phys. 5 (1964) 1252–1254.DOI
- [46]M. Gell-Mann, Symmetry violation in hadron physics. 1969.
- [47]P. Carruthers, “Broken scale invariance in particle physics,” Phys. Rept. 1 (1971) 1–29.DOI
- [48]L. Del Debbio and R. Zwicky, “Dilaton and massive hadrons in a conformal phase,” JHEP 08 (2022) 007, arXiv:2112.11363 [hep-ph].DOI
- [49]M. L. Goldberger and S. B. Treiman, “Decay of the pi meson,” Phys. Rev. 110 (1958) 1178–1184.DOI
- [50]C. Cebulla, K. Goeke, J. Ossmann, and P. Schweitzer, “The Nucleon form-factors of the energy momentum tensor in the Skyrme model,” Nucl. Phys. A 794 (2007) 87–114, arXiv:hep-ph/0703025.
- [51]M. V. Polyakov and P. Schweitzer, “Forces inside hadrons: pressure, surface tension, mechanical radius, and all that,” Int. J. Mod. Phys. A 33 no. 26, (2018) 1830025, arXiv:1805.06596 [hep-ph].DOI
- [52]G. S. Adkins and C. R. Nappi, “The Skyrme Model with Pion Masses,” Nucl. Phys. B 233 (1984) 109–115.DOI
- [53]B. Lucini and M. Panero, “SU(N) gauge theories at large N,” Phys. Rept. 526 (2013) 93–163, arXiv:1210.4997 [hep-th].DOI
- [54]P. Hernández and F. Romero-López, “The large N_c limit of QCD on the lattice,” Eur. Phys. J. A 57 no. 2, (2021) 52, arXiv:2012.03331 [hep-lat].DOI
- [55]Flavour Lattice Averaging Group (FLAG) Collaboration, Y. Aoki et al., “FLAG review 2024,” Phys. Rev. D 113 no. 1, (2026) 014508, arXiv:2411.04268 [hep-lat].
- [56]F. Alvarado and L. Alvarez-Ruso, “Extraction of the nucleon axial form factor from Lattice QCD using NNLO chiral perturbation theory,” arXiv:2606.30834 [hep-ph].
- [57]Flavour Lattice Averaging Group (FLAG) Collaboration, Y. Aoki et al., “FLAG Review 2021,” Eur. Phys. J. C 82 no. 10, (2022) 869, arXiv:2111.09849 [hep-lat].
- [58]M. Hoferichter and J. Ruiz de Elvira, “Nucleon mass: trace anomaly and σ-terms,” arXiv:2506.23902 [hep-ph].
- [59]C. Adam, C. Naya, and T. Winyard, Skyrmions coupled to the sigma meson.
- [60]C. Adam, J. Sanchez-Guillen, and A. Wereszczynski, “A BPS Skyrme model and baryons at large N_c,” Phys. Rev. D 82 (2010) 085015, arXiv:1007.1567 [hep-th].DOI
- [61]F. Meier and H. Walliser, “Quantum corrections to baryon properties in chiral soliton models,” Phys. Rept. 289 (1997) 383–450, arXiv:hep-ph/9602359.DOI
- [62]J.-L. Gervais and B. Sakita, “Large N QCD Baryon Dynamics: Exact Results from Its Relation to the Static Strong Coupling Theory,” Phys. Rev. Lett. 52 (1984) 87.
- [63]R. F. Dashen, E. E. Jenkins, and A. V. Manohar, “Spin flavor structure of large N(c) baryons,” Phys. Rev. D 51 (1995) 3697–3727, arXiv:hep-ph/9411234.DOI
- [64]A. V. Manohar, “Large N QCD,” in Les Houches Summer School in Theoretical Physics, Session 68: Probing the Standard Model of Particle Interactions, pp. 1091–1169. 2, 1998. arXiv:hep-ph/9802419.DOI
- [65]M. G. Pérez, A. González-Arroyo, and M. Okawa, “Meson spectrum in the large N limit,” JHEP 04 (2021) 230, arXiv:2011.13061 [hep-lat].
- [66]C. Bonanno, M. García Pérez, A. González-Arroyo, K.-I. Ishikawa, and M. Okawa, “Non-perturbative determination of meson masses and low-energy constants in large-N QCD,” JHEP 12 (2025) 096, arXiv:2508.05446 [hep-lat].arxiv.org/abs/2508.05446
- [67]T. A. DeGrand and E. Wickenden, “Lattice study of the chiral properties of large-N_c QCD,” Phys. Rev. D 108 no. 9, (2023) 094516, arXiv:2309.12270 [hep-lat].arxiv.org/abs/2309.12270
- [68]T. DeGrand, “Lattice calculations of the spectroscopy of baryons with broken flavor SU(3) symmetry and 3, 5, or 7 colors,” Phys. Rev. D 89 no. 1, (2014) 014506, arXiv:1308.4114 [hep-lat].DOI
- [69]T. DeGrand and Y. Liu, “Lattice study of large N_c QCD,” Phys. Rev. D 94 no. 3, (2016) 034506, arXiv:1606.01277 [hep-lat]. [Erratum: Phys.Rev.D 95, 019902 (2017)].DOI
- [70]R. Cordoba, Y. He, M. Kruczenski, D. Ping, J. Mohanty, and E. Alegria, “Gauge Theory Bootstrap with varying N_c,” arXiv:2611.16824 [hep-th].arxiv.org/abs/2611.16824
- [71]Y. Chen, Y. Li, C. Lorcé, and Q. Wang, “Nucleon axial radius,” Phys. Rev. D 110 no. 9, (2024) L091503, arXiv:2405.12943 [hep-ph].arxiv.org/abs/2405.12943
- [72]E. Braaten, S.-M. Tse, and C. Willcox, “Electroweak Form-factors of the Skyrmion,” Phys. Rev. D 34 (1986) 1482.pubmed.ncbi.nlm.nih.gov/9957310
- [73]R. Fletcher, Practical Methods of Optimization. John Wiley & Sons, Chichester, UK, 2nd ed., 2000.

